Quantifying uncertainty in decadal ocean heat uptake due ...Bablu Sinha1, Alex Megann1, Thierry Penduff2, Jean-Marc Molines2, and Sybren Drijfhout3,4,5 Fig 3:extreme heat uptake differences

(1) Global surface warming since 1850 consisted of a series of decadal timescale slowdowns (hiatus periods) followed bysurges. These are likely due to a combination of external forcing and intrinsic variability (either in the ocean alone or in thecoupled system).

(3) Composite analysis of the ensemble members with the most extreme individualdecadal heat uptake changes pinpoints the southern and northern high latitudes as theregions where intrinsic variability plays a large role: tropical heat uptake change islargely fixed by the surface forcing. Note the asymmetry between high and lowextremes.

Here we evaluate the influence of ocean intrinsic variability on global ocean heatuptake and hence the rate of global surface warming, using a 50-member ensemble ofeddy-permitting ocean general circulation model simulations (OCCIPUT ensemble),forced with identical surface atmospheric conditions during 1960-2015 (Bessières etal., 2016) .

Cumulativeoceanhe

atuptake

diffe

rence(W

m-2)

Latitude °N

(2) Air-sea heat flux, integrated zonally and accumulated with latitude provides auseful measure of ocean heat uptake (Drijfhout et al., 2014). Main diagnostic is thedifference between the 2000s and the 1990s.

Fig 2:heat uptake difference between2000-2009 (hiatus) and 1990-1999 (surge).

green=ensemble meanred=high heat uptakeblue=low heat uptake

~0.32 W m-2

~0.39 W m-2

~0.28 W m-2

OCCIPUT suggests the 2000s saw ocean heat uptake of +0.32 W m-2 compared to the1990s shared between the tropics and the high latitudes. Intrinsic ocean variabilitymodifies the mean ocean heat uptake change by up to 0.05 W m-2 or ±15%.

Anom

alou

socean

heatu

ptake

diffe

rence(W

m-2)

Latitude °N

red=4 highest wrt ens. mean (4) The western boundary currents andthe Antarctic Circumpolar Current (eddyrich regions) are responsible for therange of simulated ocean heat uptake,with the North Pacific exhibiting aparticularly strong signal.

KuroshioGulf Stream/NAC

AgulhasE. AustralianFalklands

ACC

GIN/Barents

ACCACC

Fig4:Changeinair-seaheatfluxdifference(2000sminus1990s)4 highestminus4 lowest(~5° spatialsmoothing)

W m-2

blue=4 lowest

QuantifyinguncertaintyindecadaloceanheatuptakeduetointrinsicoceanvariabilityBabluSinha1,AlexMegann1,ThierryPenduff2,Jean-MarcMolines2,andSybrenDrijfhout3,4,5

Fig 3:extreme heat uptake differences

Southern Ocean

N. high latitudes

Southern Ocean

N. high latitudes

Tropics

1NationalOceanographyCentre,UnitedKingdom2InstitutdesGéosciences del’Environnement,France3UtrechtUniversity,TheNetherlands4UniversityofSouthampton,UnitedKingdom5RoyalNetherlandsMeteorologicalInstitute,TheNetherlands

Cumulativeoceanhe

atuptake

diffe

rence(W

m-2)

Fig 5:North Pacific uptake difference between2000-2009 (hiatus) and 1990-1999 (surge).

~0.04 W m-2

~0.08 W m-2

~-0.04 W m-2

green=ensemble meanred=high heat uptake (same members as in Fig 2)blue=low heat uptake (same members as in Fig 2)

(5) Focusing on the North Pacific wemodify our diagnostic, accumulatingair-sea heat flux from 18°N within theboundaries of the North Pacific only(normalized by global ocean surfacearea). OCCIPUT suggests the 2000ssaw North Pacific ocean heat uptakeof +0.04 W m-2 compared to the1990s. Intrinsic ocean variabilitymodifies the mean ocean heat uptakechange by -0.08 to +0.04 W m-2 – oneto two times the ensemble mean value.

(7) We are looking at the influence of intrinsic variability (difference between ensemble members) on decadalvariability (difference between decades). Accordingly we split Fig 6 into four contributions: the HIGH/LOW heatuptake in each decade. Apart from HIGH in the 2000s (more on this later), the other three contributions are alldipole-like with different amplitudes, so the dipole like behaviour in Fig 6 originates from a dipole-like modepresent in each decade. It looks like the model has an intrinsic dipole-like mode of decadal variability in thePacific similar to that observed (e. g. Schmeits and Dijkstra (2002).

(8) if we separate the heat uptake curves for each decade (heat uptake minus the ensemble mean for each decade) we see that the heat uptake distribution in each decade isskewed with most values small and positive. By contrast there are a few values which are large and negative. This is the cause of the asymmetry in Fig 5 with respect toextremes of heat uptake. Members with values close to the ensemble mean can only make large excursions to negative values, not positive ones. Paradoxically, themembers labelled HIGH have negative values because HIGH refers to the decadal difference (2000s minus 1990s) which is positive, a similar argument holds for LOW.

Fig 6: Decadal difference in air sea heat flux:average HIGH minus average LOW

W m-2

Fig 7: Range within ensemble = [ HIGH (2000s) - HIGH (1990s) ] - [ LOW (2000s) - LOW (1990s) ] of decadal heat flux change

=[ ]] -- -

Latitude°N

Longitude°E

Latitud

e°N

[W m-2

W m

-2

Latitude°N

Fig 8:Components of the heatuptake difference.

red=HIGH decadal heat uptake difference - members as in Fig 5

red=HIGH decadal heat uptake difference - members as in Fig 5

blue=LOW decadal heat uptake difference - members as in Fig 5

blue=LOW decadal heat uptake difference - members as in Fig 5

Heatuptake

Heatuptake

Heatuptake

Heatuptake

Heatuptake

Heatuptake

(6) Associated spatial patterns in heatuptake show a North-South dipole likepattern. Similar patterns in SST andSSH indicate a link to the strength ofthe Kuroshio current and its extension.

2000s 1990s 2000s 1990s

2000sHIGH

2000sLOW

1990sHIGH

1990sLOW

(9) The dipole pattern becomes particularly clear if we selectmembers to maximize and minimize the heat uptake (withrespect to the ensemble mean of each decade) in bothdecades. This is the “pure” dipole mode and the examples inFig 7 are combinations of these. The magnitude of the dipoleis strongly asymmetrical – extreme negative heat uptake ismuch larger than extreme positive.

Conclusions

• OCCIPUT suggests the 2000s saw ocean heat uptake of +0.32 W m-2 compared to the1990s shared between the tropics and the high latitudes. Intrinsic ocean variabilitymodifies the mean ocean heat uptake change by up to 0.05 W m-2 or ±15%.

• The western boundary currents and the Antarctic Circumpolar Current (eddy richregions) are responsible for the range of simulated ocean heat uptake, with the NorthPacific exhibiting a particularly strong signal.

• In the North Pacific, the spatial patterns associated with extreme heat uptakedifferences show a dipole like pattern. Similar patterns in SST and SSH indicate alink to the strength of the Kuroshio current and its extension, similar to thosepreviously observed.

• There is an asymmetry in the distribution of the pattern strength, with most stateshaving weak positive amplitude, and a smaller number having large negativeamplitude. The distribution, including the asymmetry remains similar from decade todecade.

• The impact of intrinsic variability on ocean heat uptake thus reflects “allowed” or“physically consistent” transitions between states in decadal timescales. Thesetransitions are likely driven by mesoscale eddy-mean flow interaction and themechanism requires further study.

• Since it is unlikely that intrinsic variability is coordinated across the world ocean it ispossible that on rare occasions ocean heat uptake changes due to intrinsic oceanvariability could exceed the values seen in the 50-member OCCIPUT ensemble.

• The large ensemble approach is a key element in this study.

Fig 9 :The “pure” dipole: extremes of ocean heat uptake in the 1990-2000sW m-2

Longitude°E

Latitud

e°N

W m-2

Longitude°E

Latitud

e°N

Latitude°N

Latitude°N

Heatuptake

Heatuptake

cyan=strongest heat uptake associated with N. Pacific dipole

magenta=weakest heat uptake associated with N. Pacific dipole

ReferencesBessières,L.,Leroux,S.,Brankart,J.-M.,Molines,J.-M.,Moine,M.-P.,Bouttier,P.-A.,Penduff,T.,Terray,L.,Barnier,B.,andSérazin,G.,2017:DevelopmentofaprobabilisticoceanmodellingsystembasedonNEMO3.5:applicationateddyingresolution,Geosci.ModelDev.,10,1091-1106,doi:10.5194/gmd-10-1091-2017.Drijfhout,S.S.,Blaker,A.T.,Josey,S.A.,Nurser,A.J.G.,Sinha,B.andBalmaseda,M.A.,2014: Surfacewarminghiatuscausedbyincreasedheatuptakeacrossmultipleoceanbasins. Geophys.Res.Lett.,41 (22).7868-7874.https://doi.org/10.1002/2014GL061456Risbey,J.,2015:Freeandforcedclimatevariations. Nature 517, 562–563.https://doi.org/10.1038/517562aSchmeits,M.J.,andDijkstra,H.A.,2000:BimodalBehavior oftheKuroshioandtheGulfStream.J.ofPhys.Oceanogr.,31,3435-3456.https://doi.org/10.1175/1520-0485(2001)031<3435:BBOTKA>2.0.CO;2